TFS-18696; No of Pages 14

Technological Forecasting & Social Change xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Technological Forecasting & Social Change

How augmented reality apps are accepted by consumers: A comparative analysis using
scales and opinions
Alexandra Rese a,⁎, Daniel Baier a, Andreas Geyer-Schulz b, Stefanie Schreiber c
a
University of Bayreuth, Faculty of Law, Business and Economics, Chair of Innovation and Dialogue Marketing, D-95440 Bayreuth, Germany
b
Karlsruhe Institute of Technology (KIT), Institute for Information Systems and Marketing, Chair of Information Services and Electronic Markets, Kaiserstr. 12, D-76131 Karlsruhe, Germany
c
Brandenburg University of Technology Cottbus-Senftenberg, Chair of Marketing and Innovation Management, Erich-Weinert-Str. 1, D-03046 Cottbus, Germany

a r t i c l e i n f o a b s t r a c t

Article history: Increasingly, retailers rely on interactive technologies to improve consumers' shopping experiences. On the one
Received 7 September 2015 side, interactive kiosks and smart mirrors make use of dedicated devices and software to explain, configure, and
Received in revised form 12 October 2016 recommend products. On the other side, computer programs – so-called apps – are installed on the consumer's
Accepted 14 October 2016
own device for the same purpose. They can be used at home, or – if installed on a mobile device – in retail outlets
Available online xxxx
or on the move. In all cases, augmented reality (AR) can support these purposes by placing virtual content
Keywords:
(e.g. new furniture) in a real environment (the consumer's home). The overall perception and acceptance toward
Augmented reality (AR) such interactive technologies are discussed in this paper. Users' perceptions and experiences are measured by
Scan function applying a modified technology acceptance model (TAM). Four experiments, two with marker-based and two
Virtual mirror with markerless AR apps are presented to support the generalization of the results, the measurement models
E-commerce and the measurement approach. The results are satisfactory with regard to the robustness of the TAM model.
Technology acceptance model (TAM) However, the relative importance of hedonic (enjoyment, pleasure, fun) and utilitarian (information) aspects
Markerless varies for different kinds of AR apps and has to be considered for improvement to occur. From a measurement
Marker-based
point of view the acquiescence bias has to be dealt with when developing scale items.
© 2016 Elsevier Inc. All rights reserved.

1. Introduction Research has concentrated on getting AR technology to work by

Augmented reality (AR) integrates computer-generated objects
with the real environment and allows real-time interactions (Azuma,
1997). AR is rapidly gaining attention worldwide. Applications (apps)
were first developed in the 1990s, e.g. an aircraft wire bundle assembly
guidance system supporting manufacturing and repair for Boeing
(Caudell and Mizell, 1992; for more examples see Azuma, 1997).
For some time, large and unwieldy dedicated devices hampered AR
popularity. However, with the widespread adoption of smartphones
and other handheld devices the interest of developers and companies
in AR has significantly increased. Many companies are now developing
and implementing AR. Consequently, Daponte et al. (2014, p.54) state
that AR is moving from the laboratory into consumer markets. This
also applies to the retailing industry where smart or virtual mirrors for
consumer experiences were AR front-runners (Demirkan and Spohrer,
2014; Pantano and Naccarato, 2010). Pantano (2014, p. 348) empha-
sizes the potential of AR in terms of “capturing consumers' attention
and influencing their purchasing decision”.
⁎ Corresponding author.
E-mail addresses: alexandra.rese@uni-bayreuth.de (A. Rese),
daniel.baier@uni-bayreuth.de (D. Baier), andreas.geyer-schulz@kit.edu (A. Geyer-Schulz),
stefanie.schreiber@b-tu.de (S. Schreiber).
focusing on enabling technologies such as displays, other interface
devices, or algorithms. However, experiments with users have to be
made to evaluate AR devices and AR apps. So far research has focused
on the evaluation of technical demonstrators, user perception and
cognition, user task performance, or the development of user-oriented
guidelines for design (Olsson et al., 2013, p.389). Besides visualization
and interaction the overall perception and acceptance of AR is impor-
tant. Overall, user studies are scarce, and this applies in particular to
AR apps relying on handheld devices (Bulearca and Tamarjan, 2010;
Dey and Sandor, 2014).
With regard to potential usage or adoption of technical innovations
in retailing by consumers, the theoretical considerations as well as
empirical studies have extensively focused on the technology accep-
tance model (TAM) (Pantano, 2014). The TAM model has been
described “as the most influential and commonly employed theory of
information systems” (Lee et al., 2003, p.752). Originally developed as
a simple model, and relying on four basic constructs (perceived ease of
use (PEOU), perceived usefulness (PU), attitude toward using (AT) and
behavioral intention to use (BI)) (Davis, 1986, 1989), it has been
extended depending on the context with various (external) variables
(for an overview with regard to retailing see Pantano and Di Pietro,
2012). One aim of the paper is to identify potential interesting variables
with regard to AR in retailing taking previous TAM studies, but also

http://dx.doi.org/10.1016/j.techfore.2016.10.010
0040-1625/© 2016 Elsevier Inc. All rights reserved.

Please cite this article as: Rese, A., et al., How augmented reality apps are accepted by consumers: A comparative analysis using scales and
opinions, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.10.010
2 A. Rese et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx
studies from contextually similar areas and applications (e.g. virtual
worlds, innovative retailing), into account. Two main features of AR
and potential variables of interest mentioned in the literature are the
provision of additional information on products, services or the store
shopping environment and “stimulating and pleasant experiences”
(Olsson et al., 2013, p.287).
Despite its popularity, TAM has been criticized with regard to
measure validation and measurements “resulting in conflicting and
somewhat confusing findings which vary considerably in terms of
statistical significance, direction and magnitude” (Yousafzai et al.,
2007a, p.1746). In particular, the usage of self-reported measurement
scales is referred to as “the most commonly reported limitation” (Lee
et al., 2003, p.762). Even though self-reported constructs significantly
correlate with self-reported usage (e.g. Davis, 1989), there is an absence
of absolute measures making the accurate comparison of usage levels
more difficult, e.g. with regard to the technological context (Segars
and Grover, 1993). Several authors have emphasized the need for and
conducted replication studies to determine the predicted relationships
of the TAM model (for an overview including the aspect of replication
studies to validate the TAM model and measurements see Lee et al.,
2003).
This paper aims to shed light on these two gaps by investigating the
acceptance of four AR apps in the German marketing and retail markets
and comparing two different interaction principles. The mobile IKEA
catalogue app and the mobile AUTO BILD app both rely on a scan
function used in combination with a printed version of a catalogue or
magazine, whereas the Mister Spex and Ray Ban (eyewear) AR apps
implement a virtual mirror. User acceptance of these four AR apps in
consumer markets is examined with usability experiments in a labora-
tory setting. A TAM analysis is performed. Besides a scale-based
approach relying on batteries of TAM items, a more direct measurement
approach based on pre-defined adjective semantic differential pairs is
used. The results of the semantic differential and the TAM scales
approach for the four AR apps are compared with regard to the quality
of measurement and the variables included in the model.
The paper is structured as follows: The next section gives insight into
the conceptual and theoretical background for TAM and for AR with a
focus on marketing and retailing. In the following sections a modified
TAM for AR apps is developed and tested in the context of four
AR apps. In the last section, theoretical and practical implications,
limitations, and avenues for further research are presented.
2. The concept of AR against the background of marketing and
retailing
In the literature the definition of Azuma (1997, p.355) is
predominantly used; this defines AR as a system “in which 3D virtual
objects are integrated into a 3D real environment in real time”. The
superimposition of 3D virtual objects in the environment of the user
AR “enriches the sensorial perception of a person” (Daponte et al.,
2014, p. 53). AR is based on techniques developed for virtual reality
(Azuma, 1997) “but does not replace the real environment, rather AR
uses the real environment as a background” (Fonseca et al., 2014,
p.435).
For user interface technology (Rekimoto and Nagao, 1995), aspects
such as visualization and real time rendering, wireless communication,
interaction technologies, tracking and registration, data storage and
access technologies are important (Adhani and Rambli, 2012; Mekni
and Lemieux, 2014). AR enabling technologies in particular include
“(d)isplays, trackers, and graphics computers and software” (Van
Krevelen and Poelman, 2010, p.2). Recent advances in processor
performance, display technology and device equipment (e.g. video
camera, internet connection bandwidth (by LTE – long-term evolution),
GPS, and sensors such as an Inertial-Measurement Unit) have also
increased the interest in AR on mobile devices (smartphones, tablets,
glasses) (see Daponte et al., 2014). Several AR apps have been
Please cite this article as: Rese, A., et al., How augmented reality apps are accepted by consumers: A comparative analysis using scales and
opinions, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.10.010
introduced for computers and/or smartphones (for recent surveys on
the use of the term and system development see Daponte et al., 2014,
Mekni and Lemieux, 2014, or Van Krevelen and Poelman, 2010). Besides
application fields such as entertainment and games, cultural heritage,
medical, education and training, navigation and tourism, social
networking, marketing has been identified as a potential application
field (Adhani and Rambli, 2012; Gervautz and Schmalstieg, 2012;
Mekni and Lemieux, 2014). Gervautz and Schmalstieg (2012, p.30)
describe marketing together with advertising and sales as the “largest
application opportunity for AR”. Retailers can use AR to engage
customers in particular by “virtual trial and product education” as well
as gamification to enhance customer experience (Baier et al., 2015).
By presenting additional product information in terms of virtual content
the AR app can support consumers in their product decision (Adhani
and Rambli, 2012). This is in particular the case if consumers can use
their own image to interact with the product and “dress their virtual
model with the items they prefer” (Blazquez, 2014, p.100). In addition,
the experience of the customer either in the store or at home can be
enhanced, since it is fun to use the AR app interactively (Gervautz and
Schmalstieg, 2012). Other marketing application possibilities – which
are not the focus of this paper – are navigation aids to localize stores
and in-store-navigation, warehouse space optimization, brand recogni-
tion and promotion, or support of sales team members (Gervautz
and Schmalstieg, 2012). These recent surveys of AR apps indicate a
broadening of the definition of AR: AR in the wide sense is the integra-
tion of virtual objects (of any type) into a real environment in real time.
3. Theory building and research model
3.1. Technology adoption of AR in marketing and retailing: theories and
shortcomings
Fifty years after Ivan Sutherland's Sketch Pad graphical user interface
(Sutherland, 1964), research in (end user) technology acceptance is one
of the mature areas in information systems research for the simple
reason that consumer acceptance is crucial for the market success of a
new technology. And within this research area, the technology accep-
tance model (TAM) of Davis (1986) (in one of its variants) is the most
prominent model. This is supported by a Google Scholar search (June
15th, 2016) resulting in about 1.66 million links for the TAM model.
To predict technology acceptance Davis (1986) relied on a simple
Stimulus-Organism-Response model which was refined with the
Theory of Reasoned Action (TRA) of Fishbein and Ajzen (1975). Davis
(1986) suggested that the motivation of a user to use the system is
best explained by the latent construct attitude toward using (AT) – and
AT from perceived usefulness (PU) and perceived ease of use (PEOU).
Meta analyses have shown that TAM explains about 40% of the variance
in the behavioral intention to use (BI) an innovative IT system (Legris
et al., 2003). In addition, the BI of an innovative system is a direct
predictor of actual system usage explaining about 37% of the variance
after post-training in a voluntary setting and 35% of the variance in a
mandatory setting (Venkatesh et al., 2003).
TAM as a theoretical framework has been criticized due to its
simplicity for turning the research focus away from “design- and
implementation based antecedents” as well as human behavior and
experience (Benbasat and Barki, 2007, p.212). Thus, the TAM model
has been extended with various constructs, e.g. the TAM2 model
(Venkatesh and Davis, 2000) adds antecedents to PU and the TAM3
model (Venkatesh and Bala, 2008) to PEOU. Aiming at specific guidance
for designing IT systems, Wixom and Todd (2005) combine user
satisfaction models with the TAM model and rely on determinants
such as System Quality and Information Quality. Aspects of concepts
like user experience (Hassenzahl and Tractinsky, 2006) investigating
the drivers of positive experience of interactive computer technology,
e.g. the internal state of the user, characteristics of the system
(e.g. usability), and the environment, have been integrated into the
 A. Rese et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx
TAM. This also holds for pre-adoption research taking attitudes and
perceptions before behavioral intention and behavior as the focus
(Hameed et al., 2012; Karahanna et al., 1999). Overall, TAM has been
applied to analyzing various technologies (see Chuttur, 2009) with
study participants only interacting very briefly with the new system
(Davis, 1989). However, when it comes to choosing external variables
Legris et al. (2003, p.196) found “that there is no clear pattern”.
Several meta-analyses have investigated the proposed relationships
in the TAM model, which all could be positively confirmed for the most
part, but also showed inconsistencies (Legris et al., 2003; Yousafzai
et al., 2007a) or a varying relative importance of perceived usefulness
and perceived ease of use over time (Yousafzai et al., 2007a). This circum-
stance supports the undertaking of comparison studies that rely on the
same measurement instruments. From a methodological point of view
the indirect assessment and usage of self-reported item scales to
measure TAM constructs and actual system usage is criticized most
often (Chuttur, 2009). Measurement scales for PU and PEOU have
been developed and tested in Davis (1989). While Davis (1989)
proposed six items for both constructs, the number of items used varies
across studies (four items are used most frequently, e.g. Legris et al.,
2003) making a comparison with regard to the level of construct values
difficult. Hence, different data sets should be used to determine the
levels of usage (Subramanian, 1994). In addition, although the measure-
ment instruments of PU and PEOU have been confirmed to be robust in
numerous studies (Yousafzai et al., 2007a), Subramanian (1994, p.872)
found minor item scale changes to be necessary depending on e.g. the
technological context.
Davis (1989) measured AT with a more direct assessment using a
scale by Ajzen and Fishbein (1980). The scale relies on five bipolar
adjectives measured on a seven-point rating scale with a neutral
category in the middle, i.e. a semantic differential scale (see e.g. Osgood
et al., 1957). Besides displaying high measurement quality in terms of
reliability and validity, a bipolar adjective scale is described to be
“easy to administer and score, takes only a few minutes to complete,
and is applicable to a wide range of objects” (Zaichkowsky, 1985,
p. 342). The two rating approaches are both used in the paper to
measure the constructs of the TAM model.
3.2. Technology adoption of AR innovations in marketing and retailing:
determinants
With regard to user acceptance IT systems have been differentiated
into hedonic and utilitarian systems (Van der Heijden, 2004). This
duality has been transferred to interactive IT systems within the concept
of user experience (Hassenzahl et al., 2010). Utilitarian features aim to
support the users in their performance, e.g. in the fulfillment of
“particular ‘do-goals’” (Hassenzahl et al., 2010, p. 357). The functionality
of the system should be the focus and the design not distractive (Van
der Heijden, 2004). Investigating user expectations and system
requirements of AR apps relying on semi-structured interviews with
28 participants in shopping centers, Olsson et al. (2013, p. 293) found
that the value of an AR app was seen “in providing useful additional
information related to places, people, public transportation or basically
any momentarily relevant issues nearby”.
Hedonic features aim to provide the experience of fun or pleasure to
the user interacting with the IT system supporting “so-called ‘be-goals’
(e.g., to be admired, to be stimulated)” (Hassenzahl et al., 2010,
p. 357). Van der Heijden (2004, p. 696) lists for example “hedonic
content, animated images, a focus on colors, sounds, and esthetically
appealing visual layouts”. With regard to marketing and retailing
gamification and aesthetic design are important aspects (Pousttchi
and Hufenbach, 2014). Perceived enjoyment (PE) has been examined as
another important attribute of TAM. This concept has been defined as
“the extent to which the activity of using the computer is perceived to
be enjoyable in its own right, apart from any performance consequences
that may be anticipated” (Davis et al., 1992, p. 1113). In the study by
Please cite this article as: Rese, A., et al., How augmented reality apps are accepted by consumers: A comparative analysis using scales and
opinions, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.10.010
3
Olsson et al. (2013, p.296) AR apps were “expected to offer playful and
entertaining momentary experiences” (italics in original), but “artistic
and amusing content” should be avoided.
Since AR apps in marketing and retailing have been introduced
only recently, the TAM model is a suitable approach to explain
acceptance and future usage intentions. With several key words
such as “augmented reality” or “virtual worlds” and “retailing” and
“technology acceptance model” or “consumer acceptance”, Google
Scholar was used to find 14 studies (see Table 1). Most of them applied
a hypothesized research model involving AR apps and related fields
such as “virtual worlds”. With regard to the hedonic aspects of the
respective app most studies included “perceived enjoyment” or
“playfulness” as an investigated element. In contrast, focusing on
utilitarian aspects of AR apps concentrating on content being
augmented with functional information and reflected in constructs
such as perceived informativeness (PI) or information quality was quite
rare. Most studies relied on students to be participants using the app
and/or taking part in experiments. “Real” consumers moving in a retail
environment were rarely studied and surveyed. In addition, the studies
concentrated on a single app with the exception of Stoyanova et al.
(2015) and Javornik (2016a).
As a consequence of recent research on the acceptance of AR apps as
shown in Table 1 and in line with Rauschnabel and Ro (2016, p. 128) it
seems reasonable to use a basic – “quite clear and structured” – TAM
model and to extend it with two selected, “context specific” factors for
AR, namely PI and PE.
3.3. Hypotheses development
In the TAM model behavioral intention to use (BI) a technological
innovation is hypothesized to be influenced by the attitude toward
using (AT) the innovative system as well as its perceived usefulness
(PU). In turn, PU and perceived ease of use (PEOU) are considered to be
major determinants of the attitude of the user (Davis, 1989; Davis
et al., 1989). However, PU is suggested to have an additional direct
positive effect on BI without forming an attitude about the innovation,
because the users can improve their performance. In addition, PEOU is
considered to have a positive effect on PU. Several studies have
confirmed the proposed relationships and that the model performs
well (see Legris et al., 2003). We, therefore, hypothesize that the TAM
model is useful to predict usage intentions for AR apps:
Hypothesis 1. The relationships of the TAM model are valid for the four AR
apps tested. (These are the arcs labeled H1+ in Fig. 1: a) PEOU-PU, b) PU-
AT, c) PEOU-AT, d) PU-BI, and e) AT-BI.)
In the original TAM model external variables are already considered
to indirectly influence the intentions of the users by having an impact
on PEOU and PU. However, in their meta-analysis Legris et al. (2003,
p. 196) could find “no clear pattern with respect to the choice of the
external variables considered”. Enjoyment from system use and
informativeness have been identified as two major factors influencing
the attitude of users with regard to electronic and mobile commerce
(Chen and Tan, 2004; Chen and Wells, 1999; Hausman and Siepke,
2009). Perceived enjoyment (PE) has been introduced as a major behav-
ioral belief besides PU and PEOU to influence the attitude and finally the
intention to use a system (Childers et al., 2002; Davis et al., 1992; Van
der Heijden, 2004). It is also part of the TAM3 model (Venkatesh and
Bala, 2008). PE depends on the system and the time of usage (Sun and
Zhang, 2008). Ha and Stoel (2009) emphasize the unclear relationship
between the three behavioral beliefs. They propose and empirically
show that PU is the final determinant of the attitude to use a website
for online shopping. PE is considered as an antecedent of PU. The ability
of an innovative system to be playful and entertaining satisfies the need
of users “for escapism, diversion, aesthetic enjoyment, or emotional
release” (Ducoffe, 1996, p. 23). When intrinsically motivating users PE

opinions, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.10.010
Please cite this article as: Rese, A., et al., How augmented reality apps are accepted by consumers: A comparative analysis using scales and
Table 1
Research on the acceptance of AR apps and related fields in retailing in chronological order.
Study
Lee et al. (2006)
Lee and Chung (2008)
Oh et al. (2008)
Kim and Forsythe (2008)
Domina et al. (2012)
Pantano and Servidio (2012)
Olsson et al. (2013)
Poncin and Mimoun (2014)
Spreer and Kallweit (2014)
Huang and Liao (2015)
Stoyanova et al. (2015)
Javornik et al. (2016)
Javornik (2016a)
Kim and Hyun (2016)
App
Virtual-try on for clothes
(rotatable model)
Virtual shopping mall
VRIS prototype (preview –-
of furniture in a 3D virtual
showroom)
Virtual-try on for clothes
Virtual world (fashion
sites)
3D immersive store
Mobile location-aware
AR services
Magic mirror to try on
different costumes
Book catalogue
Virtual-try on for clothes
Shopping platform for – Personal emotions, usability and appearance of
shoes (interactive,
markerless, marker-based)
Virtual-try on for cosmetic –
Virtual-try on for glasses,
preview of furniture in a
real room
OVJET (navigation)
4
Model used Investigated constructs Data collection Method Retail Environment
TAM Hedonic & utilitarian shopping orientation, level Survey with 208 students SEM
of image interactivity technology, PU, PEOU, PE,
AT, BI
–- Convenience, PE, quality assurance, customer Survey with 102 students PLS
satisfaction
A. Rese et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx
Pleasure, arousal, product attitude, purchase Survey with 92 students
intention, choice confidence, decision-making
enjoyment, PEOU, PU, satisfaction, decision
confidence
TAM Technological anxiety, innovativeness, PEOU, Survey with 491 online SEM U.S. online shoppers completed
PU, PE, BI, post-use evaluation shoppers an online apparel shopping
simulation
Flow theory, TAM PE, perceived control, perceived concentration, Survey with 119 students SEM
PEOU, consumer innovativeness (novelty seeking,
independent judgement), intention to shop
Human-Computer-Interaction Store perception, PEOU, PU, PE, consumer Survey with 150 students SEM
model satisfaction
– User experience, central user requirements 16 interviews with 28 Semi-structured interviews 2 major shopping centers in
customers Tampere, Finland
Consumer satisfaction, Shopping value, emotions, perceived store Survey with 140 customers Regression analysis French-speaking toy brand
emotions and store atmosphere, patronage intention, satisfaction (parents accompanied by flagship store
atmospherics their children)
TAM PU, PEOU, PE, BI, information offer, information Survey with 96 customers Treatment group (n = 46), Book store, Germany
completeness control group (n = 50),
regression analysis
TAM Presence, PU, PEOU, perceived playfulness, Survey with 220 students PLS
perceived aesthetics, service excellence,
sustainable relationship behavior
Survey with 150 students Experiment: markerless group,
interface, attitude toward brand, position marker-based group, control
toward brand (purchase, recommend) group
Perceived augmentation, playfulness, Survey with 102 customers Store in European shopping mall
convenience, BI
Theory of interactive media Perceived augmentation, flow, affective Survey with 60 students Experiment: treatment group
effects responses, cognitive responses, BI (n = 30), control group
(n = 30), regression analysis
TAM System quality, information quality, service Survey with 114 students PLS
quality, telepresence, BI
 A. Rese et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx
Fig. 1. Research model.
“is supposed to increase the deliberation and thoroughness of cognitive
processing and lead to enhanced perceptions of extrinsic motivations
such as PU” (Sun and Zhang, 2008, p.1255). In a similar way, utilitarian
aspects of perceived informativeness (PI) have been considered. Relevant
and useful product information enables consumers to increase clarity
about the product and to come to a satisfactory product choice (Chen
and Tan, 2004). Wixom and Todd (2005) empirically showed that
information quality positively influences the PU of a data warehousing
software. PE and PI have also been emphasized with regard to AR
apps. AR provides additional digital information and at the same time
offers “hedonic experiences, such as self-expression, enjoyment and
stimulation” (Olsson et al., 2013, p.296). Against this background, we
propose the following research hypotheses as completing our research
model shown in Fig. 1:
Hypothesis 2. PI of AR apps has a direct positive effect on PU.
Hypothesis 3. PE of AR apps has a direct positive effect on PU.
Hypothesis 4. PI and PE have the same relative importance for all four
investigated AR apps.
With regard to the relative importance of the two key determinants
of technology acceptance - PU and PEOU – research has come to
inconsistent results. For example, Davis (1989) showed that the effect
of PEOU on BI is initially strong, but becomes weaker over time. With
regard to interface improvements Keil et al. (1995, p. 89) found that
PEOU cannot “compensate for low usefulness”. However, there are
also studies with an equal or even stronger effect than PU on technology
acceptance (see Yousafzai et al., 2007a). The setting of system usage
seems to play a role (Wu and Lederer, 2009). But even in voluntary
settings the impact of PU was stronger (Venkatesh, 2000). Overall,
Yousafzai et al. (2007b, p.299) come to the conclusion that PEOU is
“not a dominant predictor of usage and intentions in TAM”. With regard
to AR apps in marketing and retailing the influence of PU is considered
to be stronger on technology acceptance than PEOU due to the hedonic
Table 2
Sample characteristics of the examined AR apps.
Sample IKEA catalogue AUTO BILD
Data collection June 2013 June 2014
Sample size n = 275 n = 206
Gender 58.9% male 56.3% male
41.1% female 43.7% female
Age (mean value) 22.1 23.0
(18–34: 98.9%) (18–34: 96.1%)
Please cite this article as: Rese, A., et al., How augmented reality apps are accepted by consumers: A comparative analysis using scales and
opinions, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.10.010
5
and utilitarian aspects of AR discussed previously. This leads to our next
hypothesis:
Hypothesis 5. The PU of AR apps has a stronger influence on AT than
PEOU.
This research design aims to identify a minimal extension of
the basic TAM model which is adequate for testing the acceptance of
AR apps. Compared to the more comprehensive TAM2 and TAM3
extensions, minimalism reduces the per test cost for field experiments
of AR apps. With regard to the methodology for evaluating AR apps we
propose that a minimalistic and more direct data collection approach
relying on pre-defined adjective semantic differential pairs can replace
indirect, self-reported item scales leading to our last hypothesis:
Hypothesis 6. The results of the TAM model based on a direct approach
using pre-defined adjective semantic differential pairs are equivalent to
the results of self-reported item scales.
4. Research design and methodology
In order to assess the extended TAM model shown in Fig. 1 for differ-
ent AR apps, in total four separate laboratory experiments with two differ-
ent operationalizations of the latent constructs were conducted at a
German university in June 2013, June 2014 and in the period between
October 2014 and February 2015. The laboratory experiments of June
2014 studying two AR apps were conducted separately and at different
points of time, but for the predominant part with the same participants
(non-independent study, Turner et al., 2010). Each of the experiments
involved N 200 participants (see Table 2). The personal experience of the
participants with the respective AR app was collected via questionnaires.
4.1. AR apps under study
With regard to AR two different interaction approaches can be
distinguished: marker-based and markerless AR. In the first approach
Mister Spex Ray Ban
June 2014 October 2014–February 2015
n = 213 n = 284
58.7% male 47.5% male
41.3% female 52.5% female
22.5 23.5
(18–34: 98.6%) (18–34: 99.3%)
6 A. Rese et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx
the tracked reality consists of predefined objects that can be recognized
by image recognition. Typical markers are barcodes, QR codes, RFID tags,
or other hard-coded objects (e.g. logos, books and so on). After recog-
nizing the marker, a virtual 3D object can be displayed relatively
to the marker position and its position can be controlled by tracking
the position of the marker. In the second – much more difficult
approach – (“markerless”), only a vague definition of the real environ-
ment that needs to be tracked and enhanced with virtual objects in
real time (Adhani and Rambli, 2012) exists. The four AR apps studied
in the experiments of this paper fall in these two categories: The mobile
AUTO BILD app (introduced in March 2012) and the mobile IKEA
catalogue app belong to the first category. The markers to be tracked
are QR codes or the printed catalogue. The virtual mirror apps for Mister
Spex and Ray Ban belong to the second category (see Fig. 2). The real
environment to be tracked is the head of a consumer, the virtual objects
are glasses.
When the AUTO BILD icon is scanned by a smartphone or tablet the
mobile AUTO BILD app “brings the AUTO BILD magazine to life”: Videos,
engine sounds, audio books, 3D models, sliders to change car model
color, or additional images are provided for many stories in the maga-
zine. In addition, interactive elements such as appointment reminders,
prize competitions and live voting on the app can be used (Google
play, 2015). Interacting with the 2013 print catalogue, the mobile
IKEA catalogue app allows access to additional virtual content, such as
films or picture galleries. Sliders enable viewers to change and explore
furniture color. An “x-ray” function provides a look inside cupboards
or furniture compartments (Mashable, 2012).
Mr. Spex and Ray Ban make use of virtual try-on technology for
eyewear. Glasses can be tried on at home via a device's webcam, making
the visit to the traditional point of sale at an optician's shop seemingly
unnecessary. The live image of the webcam is used to put glasses right
on the user's nose after placing the head into an oval. Instructions are
given if the position has to be changed. If the position is correct, the
glasses adapt to all (but not too extreme and fast) head movements
and can be viewed from the side (Pohlmann, 2013). Online retailers
cover the costs of four or five sample frames and the final decision can
be made at home after a real try-on. As early as 2008 Ray Ban started
to run a website-based virtual mirror (Kell, 2011). In 2011 the Mister
Spex online shop introduced a virtual mirror requiring customers to
have PCs with webcams. In 2012 a mobile app version followed.
While all four AR apps increase the smart interaction between the
retailer and consumer due to visualizing and therefore enriching infor-
mation in a dynamic way, the virtual-try additionally has the potential
Fig. 2. Screens from IKEA's mobile catalogue app (Baldwin, 2012), AUTO BILD's mobile app (Google play, 2015), Mister Spex (Pohlmann, 2013) and Ray Ban (Radley, 2014).
Please cite this article as: Rese, A., et al., How augmented reality apps are accepted by consumers: A comparative analysis using scales and
opinions, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.10.010
to change the sales of glasses (Pantano and Timmermans, 2014).
Looking at the penetration of marker usage (barcodes, QR codes) in
Germany, two studies found that not a majority, but a certain number
of German smartphone owners make use of this technology. In July
2012, comScore (2012) relying on an online survey of a nationally
representative sample aged 13 and above found 18.6% or 5.1 million
Germans who had scanned a QR code with their smartphone.
Germany ranked first place among France, Germany, Italy, Spain and
the UK. In January 2013, a study by Pitney Bowes (2013) surveying
2000 Americans and 1000 Europeans, reported that German QR code
users (14%) ranked in the third place after US Americans (19%) and
the UK (15%). Most often German QR code scans resulted in accessing
(additional) product information (77.9%), followed by event
information (28.9%) (comScore, 2012). In particular QR codes placed
in magazines were scanned. This result applies to German users
(77.8%), but to other nationalities as well. Greater numbers of younger
users have tried to make use of QR codes in magazines. In Germany
this applies to 27% among the 18–24 years old and 23% among the
25–34 years old. Men are more likely to scan QR codes than women,
for example 60.5% of the U.S. American QR code users were male
(comScore, 2011). In contrast, while in Germany the interest in using
virtual try-on systems for eyeglasses is rather high (67.3%) (Fittkau
and Maaß Consulting, 2013), online sales are still low at 5.6% of the
corrective glasses (700,000 out of 12.43 Mio.) having been sold online
in 2015. While online sales have increased, i.e. 3% in 2012, yearly growth
has slowed down to 8% compared to 43% in 2013 (ZVA, 2016).
4.2. Questionnaire design
The questionnaire was developed in German for the laboratory
experiment examining the mobile IKEA catalogue app in 2013 and
later-on used in the 3 subsequent experiments in 2014. The first
operationalization of the extended TAM model relied on the item scales
that are shown in Appendix 1. They are based on the corresponding
literature (for PI, PEOU, AT and BI), but also on a word analysis of online
reviews with regard to the mobile IKEA catalogue app retrieved from
Google Play Store for PE and PU. For the development of the second
operationalization, we used adjective semantic differential pairs which
are an easy way to provide reliable data as an alternative measurement
approach within the TAM model. We relied on the adjectives of the
bipolar scale developed by Davis (1986) to measure AT. In addition, on-
line reviews were used to identify relevant items and adjective semantic
differential pairs (Dabholkar et al., 2000). They were considered as
 A. Rese et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx
critical incidents describing especially positive or negative situations
(Bitner et al., 1990; Flanagan, 1954). The 16 semantic differential pairs
used in the questionnaire and their assignment to the latent constructs
by an expert (one of the authors) are shown in Table 5. The survey in-
strument had been pre-tested with several PhD students and student
respondents who were smartphone users to ensure the clarity and com-
prehensiveness of the item scales. The questionnaire consisted of four
sections:
1. An overall opinion rating of the app (one to five stars, see Table 4)
followed by a free text to describe the app by relying on the format
of an online review including a brief summary and a longer reviewer
text.
2. A selection of about 10 adjectives from the 16 semantic differential
pairs (Table 5) which best describe the app.
3. The item scales of the TAM model being measured on a 7 point Likert
scale (with “1” = “strongly disagree” to “7” = “strongly agree”) in
Appendix 1.
4. Demographics of the participants (Table 2).
4.3. Participants of the laboratory experiments
For the most part the participants in the experiments were German
undergraduate university students. They were invited to take part in an
AR usability study in several courses and recruited on a voluntary basis.
As the penetration studies on marker usage for Germany show, our
samples match the age and gender distribution of consumers making
use of QR codes with their smartphone or tablet. However, for scale
construction and model evaluation a student sample is adequate. The
gender distribution in the samples showed a higher percentage of
males than females in three out of four samples. The average age of
the participants in the samples ranged between 22.1 and 23.5 years
(see Table 2).
4.4. Procedure of the laboratory experiments
Two of the examined AR apps – mobile IKEA catalogue app and
mobile AUTO BILD app – involved the scan of a marker with a mobile
device, such as a smartphone or tablet. In addition, a printed edition of
a catalogue or a magazine was required for the marker. The respondents
should make use of the scan function to access additional product
information. The increasing penetration of smartphones and tablets
was the reason why in contrast to the IKEA catalogue experiment in
2013, no devices were provided, but the respondents should use their
own mobile device in 2014. In Germany the number of mobile users
increased from 37.9% in the second quarter of 2013 to 48.4% in the
second quarter of 2014 (AGOF, 2013, 2014). The laboratory experiments
on the virtual mirror relied mainly on stationary devices but also on
laptops with an integrated webcam (see Table 3).
The laboratory experiments all followed the standard procedure in
usability testing (Cockton et al., 2008; Dumas and Fox, 2008): First,
the participants were informed about the AR usability study and made
familiar with the functionality and use of the respective AR app. To get
to know the AR app and the additional virtual content the participants
Table 3
Characteristics of the laboratory experiments.
Laboratory experiment IKEA catalogue
Device Same (standard) tablet
Additional material IKEA catalogue
AR technology Make use of the scan function
Task of participants Describe the additional content of several
catalogue pages, search for specific furniture
Please cite this article as: Rese, A., et al., How augmented reality apps are accepted by consumers: A comparative analysis using scales and
opinions, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.10.010
7
had to complete several small tasks which are comparable for the
IKEA/AUTO BILD and Mister Spex/Ray Ban experiments. For example,
in the Mister Spex experiment the participants should select two
eyeglass models. Non-spectacle wearers should choose two sunglass
models, spectacle wearers should choose either two sunglass models
or two eyeglass models for themselves. When the tasks were finished,
the participants filled out a questionnaire to reflect their experience
with the AR app (see Fig. 3).
5. Results
5.1. Descriptive results of opinions on the AR apps
With regard to an overall opinion rating corresponding to online
reviews (one to five stars indicating overall satisfaction) the two
marker-based AR apps (mobile IKEA catalogue app, mobile AUTO BILD
app) are very close in average rating and distribution of stars (see
Table 4). The opinion rating of the participants ranged to the main
part between 3 and 4 stars. The two markerless AR apps (Mister Spex,
Ray Ban) also follow a similar pattern in star ratings with higher average
ratings and stars distributed more in the range of 4 and 5 stars. Overall,
the star ratings were below the average star rating of 4.29 for products
with hedonic aspects reported by Pan and Zhang (2011).
With regard to the adjective semantic differential pairs, respondents
had to select at most 10 descriptive words out of 32 adjectives. To focus
the answers of the respondents on important attributes of the app,
participants were asked to provide the 10 adjectives which best
describe the app. However, this restriction was not enforced. In the
experiments on average more than six words were ticked off by the
respondents (IKEA: 6.6; AUTOBILD: 6.3; Mister Spex: 7.2; Ray Ban:
6.9). Very few cases in each data set except the IKEA sample displayed
N10 answers. The values given in Table 5 show the percentage of the
respondents of the respective experiment who had picked a certain
word. For example, of the 284 Ray Ban experiment participants the
word “good” was selected by 122 respondents and “poor” by 3. All
four AR apps were predominantly described with positive adjectives,
in particular Mister Spex. The two virtual mirror apps Mister Spex and
Ray Ban were to a large percentage characterized by utilitarian aspects
such as ‘practical’, ‘helpful’ or ‘useful’. The informational aspect was
considered as beneficial in the case of the two scan function apps.
Nevertheless, the speed and maturity level was assessed to be in need
of improvement for all four apps (see Table 5) and, in addition, consid-
erable differences (e.g. in speed between Mister Spex and Ray Ban)
were discovered. The analysis of these differences led to suggestions
for improving an app.
SmartPLS 2.0 was used to analyze the four data sets. The analysis
included reliability and validity tests, followed by the assessment of
the structural model. The examination of the measurement model
included 26 items and 6 latent constructs. Table A1 in Appendix 2
shows the results for reliability, validity and discriminant validity.
Cronbach's Alpha (CA), measuring internal consistency, exceeded the
threshold of 0.7 for all six latent constructs in all samples (Nunnally,
1978). Almost all items loaded at least with a factor loading of 0.7 on
AUTO BILD Mister Spex /Ray Ban
Smartphone/tablet of the participants Desktop PCs/laptops with webcams / tablet
of the participants
AUTO BILD –
magazine
Make use of the scan Make use of the virtual mirror
function
Describe the additional content of several magazine Select two different eyeglass models
pages, search for information on specific car models
8 A. Rese et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx
Fig. 3. a–c. Procedure of the laboratory experiment with regard to the mobile IKEA catalogue app (3a: Information on AR usage, 3b: Perform a small task (look behind cupboard doors),
3c: Fill out questionnaire).
their respective factors. The only exception are the items of the
construct “behavioral intention to use” in the case of the Mister Spex
experiment for which factor loadings are in the range of 0.618–0.860.
In addition, all composite reliability (CR) values were above the thresh-
old of 0.60 (Bagozzi and Yi, 1988). The same holds for the AVE (Average
Variance Extracted) of all constructs for a threshold of 0.50 (Fornell and
Larcker, 1981). Overall, the results indicate that all measures used in this
study are robust in terms of their composite reliability and convergent
validity in several experimental settings, indicating reliability.
5.2. Analysis of results on the acceptance of the AR apps
With regard to discriminant validity we checked whether the square
root of AVE was greater than the correlation between the constructs
(Fornell and Larcker, 1981). This criterion was met with one exception
(correlation PU-AT in the Ray Ban case) (see Table A1 in Appendix 2).
To check for multicollinearity, the variance inflation factor (VIF) was
calculated for each factor which with one exception (PU in the AUTO
BILD experiment) was lower than 5 (Hair et al., 2011). A composite
score for all key factors of the TAM model was calculated. Comparing
the self-reported item scale based constructs the mean values for
PEOU and PU were highest for all four AR apps while the lowest mean
value was for BI (see Table A2 in Appendix 2). The distribution of the
rankings of the mean values followed the same pattern in the case of
the markerless and marker-based apps. Overall, the markerless
apps showed consistently higher values in each dimension of the TAM
model.
The estimation of the TAM model by partial least squares revealed
that the relationships were confirmed for all AR apps with the exception
of PEOU (Hypothesis 1b, c, d, e). In contrast to PEOU, the impact of PU on
AT and BI was highly significant (Hypothesis 5). The impact of the util-
itarian (PI) and hedonic (PE) variables on PU was confirmed
(Hypothesis 2, 3). The models showed a substantial good fit with high
R2 values for the latent variables starting from 0.561 (R2 values of
0.33 are described as moderate and R2 values of 0.67 as substantial,
Henseler et al., 2009). The model explains over 60% of the variance of
BI for each of the AR apps (see Fig. 4). With regard to the statistical sig-
nificance of the path coefficients, the effect size of each of the predictor
latent variables showed that the effect of PI on PU was larger for the
markerless AR apps than for the two marker-based AR apps. On the
other hand, PE had a larger impact on PU for the two marker-based
Table 4
Star ratings with respect to the examined AR apps.
Sample IKEA catalogue
Share of ratings with 1 star 2.5% (n = 7)
Share of ratings with 2 stars 14.2% (n = 39)
Share of ratings with 3 stars 34.2% (n = 94)
Share of ratings with 4 stars 41.1% (n = 113)
Share of ratings with 5 stars 8.0% (n = 22)
Missing 0 0
Average rating 3.4
Please cite this article as: Rese, A., et al., How augmented reality apps are accepted by consumers: A comparative analysis using scales and
opinions, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.10.010
apps (see Table 6). Therefore, Hypothesis 4 was not confirmed. In
turn, PU had at least a moderate effect on BI (with the exception of
Ray Ban).
In addition to the scale-based approach, the adjective semantic
differential pairs were sorted into the TAM categories (see Table 5). A
combined score was calculated as follows: If the respondent answered
with a positive word, a value of 1 was added. Correspondingly, for a
negative word, a value of 1 was subtracted. The combined score was
recoded within the interval [1, 7].
Comparing the approach based on adjective pairs with the scale-
based approach, the former also reflects the basic TAM model well as
proposed by Hypothesis 6 (see Fig. 5 and Table 6). However, the
proposed positive effect of the two external variables PE and PI on PU
is not confirmed for all four samples (Hypothesis 2, 3). This is in
particular true for PI. In contrast to the scale-based approach the
PEOU–PU relationship is significant. However, with the exception of
the IKEA sample the R2 of PU, AT and BI is rather low. With regard to
the level of values of the basic TAM constructs the composite scores
relying on the adjective pairs confirmed the results of the self-
reported item scales. Again the markerless apps were rated highest in
terms of BI and PU. In particular, BI was low for the marker-based
apps despite the more favourable attitude regarding PI (see Table A3
in Appendix 2).
6. Conclusions
This study provides insights when measuring consumer/user
acceptance of AR apps in marketing and retailing. Relying on two
different measurement approaches (item scales, adjective semantic
differential pairs), the results support the validity of the basic TAM
model (Davis et al., 1989) for customer acceptance with regard to the ini-
tial adoption of a technology. The second operationalization of the TAM
model with adjective semantic differential pairs offers the advantage of
a better root cause analysis of the AR apps' specific strengths and weak-
nesses. Of course, additional work is necessary to unfold the complete
potential of this approach. The majority of the participants had used
the AR apps for the first time (initial adoption). For example, b 20% of
the participants had previously ordered eyewear on the internet (Mister
Spex: 15.5%, Ray Ban 18.8%). About 17.6% of the participants were read-
ing AUTO BILD magazine rather frequently to get information about cars
(up to 5 on a scale ranging from 1 = ‘never’ to 7 ‘frequently’).
AUTO BILD Mister Spex Ray Ban
2.4% (n = 5) 0.5% (n = 1) 0% (n = 0)
11.7% (n = 24) 3.3% (n = 7) 5.4% (n = 15)
36.4% (n = 75) 12.7% (n = 27) 25.9% (n = 72)
39.3% (n = 81) 57.3% (n = 122) 51.4% (n = 143)
10.2% (n = 21) 26.3% (n = 56) 17.3% (n = 48)
0 6
3.4 4.1 3.8
 A. Rese et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx 9

Table 5
Frequency of positive and negative descriptive adjectives.

Negative adjective (in%) IKEA cata-logue AUTO BILD Mister Spex Ray Ban Positive adjective (in%) IKEA cata-logue AUTO BILD Mister Spex Ray Ban

(PI) Not informative 13.8 9.2 1.4 3.5 Informative 36.4 57.8 16.4 27.8
(PE) Unexciting 6.2 14.1 0.9 2.1 Exciting 30.5 17.5 17.8 25.0
(PU) Impractical 16.4 10.2 2.3 5.3 Practical 34.5 36.9 59.6 62.7
(PU) Unhelpful 9.8 7.3 2.3 0.7 Helpful 29.1 26.7 54.5 60.9
(PU) Not interactive 1.8 3.4 2.3 1.1 Interactive 35.3 37.4 33.8 26.1
(PEOU) Immature 37.8 34.5 29.6 39.8 Mature 3.6 6.3 10.8 3.2
(PEOU) Confusing 9.5 1.5 2.3 4.6 Clear 31.3 46.1 35.2 34.9
(PEOU) Slow 22.2 23.3 6.1 50.4 Fast 35.3 34.5 53.1 16.9
(PEOU) Inadequate 12.7 8.7 8.0 9.2 Perfect 5.1 6.3 13.6 3.5
(PEOU) Complicated 8.4 2.9 0.9 2.5 Simple 47.6 53.9 58.2 50.7
(AT) Not innovative 4.0 9.2 2.3 3.2 Innovative 52.4 32.0 55.9 41.5
(AT) Weak 6.5 4.9 2.3 2.1 Great 5.1 5.3 20.2 9.5
(AT) Poor 3.6 2.4 0.5 1.1 Good 60.7 37.4 51.6 44.4
(BI) Not worth recommending 5.5 9.2 0.9 2.8 Worth recommending 20.7 19.4 56.8 41.5
(BI) Superfluous 24.7 23.8 1.4 1.8 Useful 37.5 23.8 67.6 69.4
(BI) Absurd 14.2 11.7 2.8 0.7 Meaningful 17.1 15.0 45.1 37.0
Number of responses 542 363 142 371 (1.3) 1267 (4.6) 940 1385 1576 (5.5)
(average number of responses (2.0) (2.0) (0.7) (2.7) (6.5)
per participant)

From a theoretical perspective, the results shed light on factors
influencing the adoption of AR in marketing and retailing. Regarding
the two different – direct and indirect – measurement approaches
most proposed relationships of the basic TAM model and its constructs
were confirmed for both interaction approaches. Markerless apps
outperformed marker-based apps with regard to the basic TAM
constructs, providing an advantage with regard to recommendation
intentions and usage intentions. These apps are characterized by
a high augmentation level that allows for a personalized spatially
dynamic augmentation experience. With regard to a future adoption it
should be noted that a rather low number of participants considered
the mobile IKEA catalogue app and the mobile AUTO BILD app to be
‘worth recommending’. In addition, these two apps are characterized
more often to be ‘superfluous’. A practical implication of the adjective
pairs used to describe the AR apps is that the speed and maturity level
of all four apps could be improved.
However, there are also differences to be considered. The results of
the adjective semantic differential pairs approach do not support the
statistically insignificant relationship between PEOU and PU and mostly
not the significant effects of PI and PE. However, the last effect can be
explained at least partially by the fact that just a single adjective
semantic differential pair was assigned to PI and PE compared to at
least three pairs for all of the other latent constructs. For the scale-
based approach the two specific aspects of AR apps (PI, PE) were
confirmed with regard to their utilitarian focus. There are differences
in the relationship of these two aspects with regard to marker-based
and markerless AR apps. The effect of PI on PU is lower with the scan
function, while the influence of PE is higher. However, when using the
combined scores of the semantic adjective pairs only the IKEA experi-
ment showed highly significant effects of PI and PE. One reason might
be the adjective pairs were identified as relying on a word analysis of
online reviews with regard to the mobile IKEA catalogue app. The
word analysis should have also been used to rely on specific content
for the other AR apps. Another reason might be that PI and PE were
only represented by one adjective pair. Davis (1986) used a five-item
semantic differential scale to measure attitude. With regard to
sensitivity, up to four adjectives to measure a dimension are considered
to be adequate (Heise, 1970). This implies that the different underlying
aspects of PI (e.g. accuracy, relevance, vividness, personalization) and PE
(e.g. gamification) should be investigated for AR apps in more detail.
Regarding the relationship between PEOU and PU, the results of the
scale-based approach emphasize the usefulness of the AR apps. With
regard to PEOU no (partial) mediating effect on AT was found. The
non-significant relationship between PU and the less significant one
between AT confirmed the findings of the meta-analysis of Lee et al.
(2003, p.759) which describe PEOU as an “unstable measure”. The reason

Fig. 4. TAM model of the four AR apps using scales and smartPLS 2.0.

Please cite this article as: Rese, A., et al., How augmented reality apps are accepted by consumers: A comparative analysis using scales and
opinions, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.10.010
10 A. Rese et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx

Table 6
Effect size of the predictor latent variables (N = 0.02 weak, N = 0.15 moderate, N = 0.35 large effect).

Effect size (f2) Scales Adjective pairs

IKEA catalogue AUTO BILD Mister Spex Ray Ban IKEA catalogue AUTO BILD Mister Spex Ray Ban

PI ➔ PU 0.255 0.277 0.542 0.380 0.136 0.012 0.021 0.003

PE ➔ PU 0.451 1.143 0.373 0.403 0.042 0.024 0.014 0.020
PEOU ➔ PU 0.003 0.017 0.008 0.105 0.042 0.039 0.081 0.043
PEOU ➔ AT 0.065 0.047 0.016 0.040 0.090 0.095 0.024 0.029
PU ➔ AT 0.977 2.164 1.519 1.895 0.162 0.105 0.049 0.020
PU ➔ BI 0.310 0.177 0.282 0.078 0.361 0.100 0.064 0.085
AT ➔ BI 0.152 0.222 0.086 0.186 0.079 0.062 0.174 0.035

could be the inherently easy-to-use nature of AR systems resulting in no
or only a small influence on PU (Subramanian, 1994). However, also the
choice of items for operationalization PEOU could be the decisive reason.
By contrast, the results of the semantic adjective pairs approach
support a significant relationship between PEOU and PU. In addition,
the direct influence of PEOU on AT is almost as strong as the one for
PU. This finding is in line with the results of the meta-analysis of
Yousafzai et al. (2007b). They found for student samples and laboratory
experiments that the effect of PEOU on AT was higher than for PU due to
the tasks being subject to a time limit. The insignificant relationship
within the scale-based approach might be the result of an acquiescence
bias implied by the composite score's low standard deviation. According
to Friborg et al. (2006) semantic-differential based scorings can reduce
acquiescence bias and display a better data fit. In this kind of response
bias respondents tend “to respond positively to items irrespective of
items contents” (Friborg et al., 2006, p. 874). This problem exists be-
cause the item scales used in the experiments contained only items for-
mulated in a positive way. The scale items predominantly used in the
four studies emphasize the underlying intuitiveness of the AR interface
requiring little cognitive effort. Negative aspects of AR app usage such as
frustration, cumbersome handling, or the app being rigid and inflexible
which are also reflected in PEOU items (Legris et al., 2003) could pro-
vide an alternative when developing a measurement instrument.
The measures used in this study indicated a high degree of reliability
being used with satisfactory validity results in several experimental
studies. However, there are a few anomalies. The correlation between
AT and PU is high and there is a problem with multicollinearity in
AUTO BILD and with discriminant validity in Ray Ban. This could consti-
tute a problem especially when the PU is a highly modified, almost new,
scale. In addition, the indicator reliability of one of the items of the BI
scale is lower than 0.70 in Mister Spex. Another anomaly is the insignif-
icant relationship between PEOU and AT in Mister Spex. The little vari-
ation of PEOU (STDEV =0.57) can be considered to be an explanation.
The quality of the measurement constructs did not vary across the
two different interaction approaches implying that item changes
might only be necessary if the technological contexts differ considerably
(Subramanian, 1994).
Altogether, this study makes three main contributions to the TAM
literature from a measurement perspective. First, the TAM model is a
robust model that is also valid for AR apps. Second, with regard to
scale development the problem of an affirmative response bias has to
be considered when developing scale items. And, last but not least,
several experiments are necessary to test the quality of scale items.
6.1. Limitations and future research
Several limitations of this study are noteworthy and shed light on
avenues for future research. First, the samples relied on students and
experimental, but not authentic, settings. This is a typical shortcoming
of university-based TAM research. However, a student sample is
adequate for the goal of model and scale evaluation and to detect usabil-
ity problems. To overcome this shortcoming and to get a more realistic
picture of AR apps used in real environments, online consumer reviews
should be included in the analysis. Control groups using an interactive,
but non-augmented app to establish the effect of AR should also be
employed. Due to the young age group of participants all AR apps
were assessed from a larger part as being ‘simple’. The result might be
different for an older age group (e.g. Chong, 2013; Gurtner et al.,
2014). In addition, the adoption of AR apps may vary depending on
the cultural context (Choi et al., 2014; Lee et al., 2015). Second, variables
(e.g. stage of decision process) could be integrated into the model acting
as moderator with regard to hedonic and functional utility (Javornik,
2016b). It may be expected that hedonic utility increases product eval-
uation when consumers are at the consideration set formation stage.
Functional utility may be increasing product evaluation when con-
sumers are at the decision-making stage. Online consumer reviews
could help to identify additional important external variables, underly-
ing aspects/dimensions of PI and PE, or valuable ideas for technical or

Fig. 5. TAM model of the four AR apps using adjective pairs and smartPLS 2.0.

Please cite this article as: Rese, A., et al., How augmented reality apps are accepted by consumers: A comparative analysis using scales and
opinions, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.10.010
 A. Rese et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx 11

design improvements. Third, this study did not consider potential inter- two different technologies (marker-based, markerless) as examples.
dependencies in the case of participants taking part in several experi- Other AR apps in other contexts such as fashion, jewelry, or toys, should
ments. Fourth, this study explored four different AR apps relying on also be explored with regard to TAM.

Appendix 1. Item scales used in the experiments (case: mobile IKEA catalogue)

Reference

Perceived enjoyment (PE)

PE1 Using the IKEA app is really fun. Online reviews
PE2 The scan function and its elements are a nice gimmick.
PE3 It is fun to discover the scan function and its elements.
PE4 The IKEA app invites you to discover the IKEA catalogue.

Perceived informativeness (PI)

PI1 The IKEA app provides detailed information about the furniture. Ahn et al. (2004); Hausman and Siepke (2009)
PI2 The IKEA app provides the complete information about the furniture.
PI3 The IKEA app provides information that helps me in my decision.
PI4 The IKEA app provides information to compare products.

Perceived ease of use (PEOU)

PEOU1 I found the IKEA app to be very easy to use. Venkatesh and Davis (2000); Gefen et al. (2003)
PEOU2 The IKEA app was intuitive to use.
PEOU3 It was easy to learn how to use the IKEA app.
PEOU4 Handling the scan function and its elements was easy.

Perceived usefulness (PU)

PU1 For me, the IKEA app has great value. Online reviews
PU2 The IKEA app provides beautiful interior design ideas.
PU3 The IKEA app is very inspiring in terms of interior design ideas.
PU4 The IKEA app is perfect for keeping the overview of the furniture.

Attitude toward using (AT)

AT1 I am positive about the IKEA app. Ahn et al. (2004); Porter and Donthu (2006)
AT2 The IKEA app is so interesting that you just want to learn more about it.
AT3 It just makes sense to use the IKEA app.
AT4 The use of the IKEA app is a good idea.
AT5 Other people should also use the IKEA app.

Behavioral intention to use (BI)

If I were to buy furniture in the future, I would…
BI1 …download or use the IKEA app immediately. Ahn et al. (2004)
BI2 …give the IKEA app priority over the printed catalogue.
BI3 …give the IKEA app priority over the catalogues of other providers.
BI4 I will recommend using the IKEA app to my friends.
BI5 I will use the IKEA app regularly in the future.

Note: All items were measured on a 7 point Likert scale anchored from 1 (strongly disagree) to 7 (strongly agree).

Appendix 2. Additional tables

Table A1
Correlations, item loadings, Average Variance Extracted (AVE), Composite Reliability (CR), Cronbach Alpha (CA), and Variance Inflation Factor (VIF) of the constructs in the samples
calculated with smartPLS 2.0.

Outer loadings of items Cross-correlations AVE N 0.50 CR N 0.60 CA N 0.70 VIF b 5

PI PE PEOU PU AT BI

IKEA
PI 0.782–0.875 0.84 0.712 0.908 0.866 1.926
PE 0.806–0.929 0.57 0.87 0.757 0.925 0.892 2.948
PEOU 0.827–0.905 0.38 0.59 0.87 0.761 0.927 0.895 1.582
PU 0.869–0.927 0.67 0.76 0.49 0.90 0.802 0.942 0.917 4.244
AT 0.868–0.931 0.62 0.70 0.52 0.78 0.90 0.815 0.957 0.943 3.176
BI 0.759–0.929 0.59 0.69 0.47 0.79 0.75 0.86 0.747 0.936 0.914 3.119

AUTO BILD
PI 0.804–0.876 0.85 0.715 0.909 0.868 1.840
PE 0.872–0.899 0.51 0.89 0.788 0.937 0.911 3.902
PEOU 0.857–0.941 0.11 0.38 0.90 0.805 0.943 0.919 1.219
PU 0.821–0.921 0.65 0.82 0.22 0.88 0.774 0.932 0.902 5.391
AT 0.860–0.913 0.61 0.80 0.30 0.84 0.89 0.785 0.948 0.932 4.853
BI 0.791–0.893 0.60 0.70 0.24 0.80 0.81 0.84 0.711 0.925 0.898 3.538

Mister Spex
PI 0.763–0.869 0.81 0.656 0.884 0.827 2.363
PE 0.770–0.900 0.54 0.83 0.693 0.900 0.853 2.343
PEOU 0.771–0.860 0.15 0.39 0.83 0.685 0.897 0.847 1.204

(continued on next page)

Please cite this article as: Rese, A., et al., How augmented reality apps are accepted by consumers: A comparative analysis using scales and
opinions, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.10.010
12 A. Rese et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx

Table A1 (continued)

Outer loadings of items Cross-correlations AVE N 0.50 CR N 0.60 CA N 0.70 VIF b 5

PI PE PEOU PU AT BI

PU 0.838–0.879 0.74 0.70 0.20 0.86 0.736 0.918 0.880 4.397

AT 0.767–0.915 0.69 0.65 0.23 0.79 0.84 0.707 0.923 0.895 3.165
BI 0.618–0.860 0.63 0.57 0.12 0.77 0.71 0.80 0.637 0.897 0.854 2.684
Ray Ban
PI 0.766–0.810 0.79 0.626 0.870 0.803 1.948
PE 0.763–0.901 0.57 0.83 0.682 0.895 0.842 2.250
PEOU 0.867–0.918 0.31 0.44 0.89 0.795 0.940 0.915 1.263
PU 0.783–0.850 0.65 0.67 0.31 0.82 0.670 0.890 0.837 3.823
AT 0.827–0.889 0.66 0.69 0.37 0.83 0.87 0.758 0.940 0.920 4.345
BI 0.706–0.874 0.56 0.57 0.28 0.73 0.76 0.82 0.670 0.910 0.876 2.525

Table A2
Mean values (standard deviation) of the TAM constructs in the different samples based on scale-based scores calculated with IBM Statistics SPSS 22.

App PI PE PEOU PU AT BI

IKEA 4.67 (1.35) 5.39 (1.37) 5.73 (1.19) 4.82 (1.46) 4.48 (1.49) 3.72 (1.72)
(Rank) (3) (3) (4) (3) (3) (3)
AUTO BILD (Rank) 4.26 (1.35) 4.89 (1.46) 6.07 (1.07) 4.33 (1.42) 4.25 (1.46) 3.44 (1.56)
(4) (4) (3) (4) (4) (4)
Mister Spex (Rank) 5.19 (1.11) 6.07 (0.87) 6.51 (0.57) 5.56 (1.14) 5.87 (1.02) 4.77 (1.32)
(2) (1) (1) (1) (1) (1)
Ray Ban 5.20 (1.07) 5.86 (0.97) 6.27 (0.87) 5.52 (1.04) 5.64 (1.12) 4.24 (1.43)
(Rank) (1) (2) (2) (2) (2) (2)

Table A3
Mean values (standard deviation) of the TAM constructs in the different samples based on semantic adjective scores calculated with IBM Statistics SPSS 22.

App PI PE PEOU PU AT BI

IKEA 4.98 (1.63) 4.73 (1.63) 4.24 (1.25) 4.71 (1.18) 4.83 (1.00) 4.40 (1.38)
(Rank) (2) (1) (3) (4) (3) (3)
AUTO BILD (Rank) 5.46 (1.96) 4.10 (1.69) 4.46 (1.06) (2) 4.80 (1.06) (3) 4.58 (0.97) (4) 4.14 (1.37) (4)
(1) (4)
Mister Spex (Rank) 4.45 (1.19) 4.51 (1.20) 4.74 (0.95) 5.41 (1.20) 5.23 (0.94) 5.64 (1.07)
(4) (3) (1) (2) (1) (1)
Ray Ban 4.73 (1.52) 4.69 (1.41) 4.02 (0.96) 5.53 (0.99) 4.89 (1.41) 5.43 (1.04)
(Rank) (3) (2) (4) (1) (2) (2)

References
Adhani, N.I., Rambli, D.R.A., 2012. A survey of mobile augmented reality applications.
Proceedings of the 1st International Conference on Future Trends in Computing and
Communication Technologies, pp. 89–96.
AGOF (Arbeitsgemeinschaft Online Forschung - Online Research Group), 2013h. Mobile
facts 2013-II retrieved September 25, 2016, https://www.agof.de/download/
Downloads_Mobile_Facts/Downloads_Mobile_Facts_2013/Downloads_Mobile_
Facts_2013_II/mf%202013-II%20AGOF_mobile_facts_2013-II.pdf?63ef58.
AGOF (Arbeitsgemeinschaft Online Forschung - Online Research Group), 2014h. Mobile
facts 2014-II retrieved September 25, 2016, https://www.agof.de/download/
Downloads_Mobile_Facts/Downloads_Mobile_Facts_2014/Downloads_Mobile_
Facts_2014_II/mf%202014-II%20AGOF%20Grafiken.pdf.
Ahn, T., Seewon, R., Han, I., 2004. The impact of the online and offline features on the
user acceptance of internet shopping malls. Electron. Commer. Res. Appl. 3 (4),
405–420.
Ajzen, I., Fishbein, M., 1980. Understanding Attitudes and Predicting Social Behaviour.
Prentice-Hall, Englewood Cliffs, N.J.
Azuma, A.T., 1997. A survey on augmented reality. Presence 6 (4), 355–385.
Bagozzi, R.P., Yi, Y., 1988. On the evaluation of structural equation models. J. Acad. Market
Sci. 16 (1), 74–94.
Baier, D., Rese, A., Schreiber, S., 2015. Analyzing online reviews to measure technology
acceptance at the point of scale — the case of IKEA. In: Pantano, E. (Ed.), Successful
Technological Integration for Competitive Advantage in Retail Settings. Hershey PA,
IGI Global, pp. 168–189.
Baldwin, R., 2012. Ikea's augmented reality catalog will let you peek inside furniture
retrieved September 25, 2016, https://www.wired.com/2012/07/ikeas-augmented-
reality-catalog-lets-you-peek-inside-the-malm/.
Benbasat, I., Barki, H., 2007. Quo vadis TAM? J. Assoc. Inf. Syst. 8 (4), 211–218.
Bitner, M.J., Booms, B., Tetreault, M., 1990. The service encounter: diagnosing favorable
and unfavorable incidents. J. Mark. 54 (1), 71–84.
Blazquez, M., 2014. Fashion shopping in multichannel retail: the role of technology in
enhancing customer experience. Int. J. Electron. Commer. 18 (4), 97–116.
Bowes, P., 2013. Getting ahead of the emerging QR code marketing trend retrieved
September 25, 2016, http://pressroom.pitneybowes.co.uk/getting-ahead-of-the-
emerging-qr-code-marketing-trend-/.
Bulearca, M., Tamarjan, D., 2010. Augmented reality: a sustainable marketing tool? Global
Bus. Manage. Res. 2 (2 & 3), 237–252.
Caudell, T.P., Mizell, D.W., 1992. Augmented reality: an application of heads-up
display technology to manual manufacturing processes. In: Shriver, B.D. (Ed.),
System Sciences, 1992 Proceedings of the Twenty-Fifth Hawaii International
Conference on. IEEE vol. 2. IEEE Computer Society press, Los Alamitos, CA,
pp. 659–669.
Chen, L.-D., Tan, J., 2004. Technology adaptation in e-commerce: key determinants of
virtual stores acceptance. Eur. Manag. J. 22 (1), 74–86.
Chen, Q., Wells, W.D., 1999. Attitude toward the site. J. Adv. Res. 39 (5), 27–37.
Childers, T.L., Carr, C.L., Peck, J., Carson, S., 2002. Hedonic and utilitarian motivations for
online retail shopping behavior. J. Retail. 77 (4), 511–535.
Choi, J., Lee, H.J., Sajjad, F., Lee, H., 2014. The influence of national culture on the at-
titude towards mobile recommender systems. Technol. Forecast. Soc. Chang. 86,
65–79.
Chong, A.Y.-L., 2013. Mobile commerce usage activities: the roles of demographic and
motivation variables. Technol. Forecast. Soc. Chang. 80 (7), 1350–1359.
Chuttur, M., 2009. Overview of the technology acceptance model: origins, developments
and future directions. Sprouts 9 (37), 1–23.
Cockton, G., Woolrych, A., Lavery, D., Sears, A., Jacko, J., Tsuchiya, I., Grandy, G., 2008.
Inspection based evaluations. In: Sears, A., Jacko, J. (Eds.), The Human-Computer Inter-
action Handbook. Fundamentals, Evolving Technologies, and Emerging Application,
Chap. 59. Associates (Taylor and Francis Group), New York, pp. 1171–1191.
comScore, 2011. 14 million Americans scanned QR codes on their mobile phones in June
2011 retrieved September 25, 2016, http://www.comscore.com/ger/Insights/Press-
Releases/2011/8/14-Million-Americans-Scanned-QR-or-Bar-Codes-on-their-Mobile-
Phones-in-June-2011.
comScore, 2012. QR code usage among European smartphone owners doubles over past
year retrieved September 25, 2016, http://www.comscore.com/ger/Insights/Press-
Releases/2012/9/QR-Code-Usage-Among-European-Smartphone-Owners-Doubles-
Over-Past-Year.

Please cite this article as: Rese, A., et al., How augmented reality apps are accepted by consumers: A comparative analysis using scales and
opinions, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.10.010
 A. Rese et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx
Dabholkar, P.A., Shepherd, C.D., Thorpe, D.I., 2000. A comprehensive framework for
service quality: an investigation of critical conceptual and measurement issues
through a longitudinal study. J. Retail. 76 (2), 139–173.
Daponte, P., De Vito, L., Picariello, F., Riccio, M., 2014. State of the art and future developments
of the augmented reality for measurement applications. Measurement 57, 53–70.
Davis, F.D., 1986. A Technology Acceptance Model for Empirically Testing New End-User
Information Systems: Theory and Results (PhD thesis) Massachusetts Institute of
Technology, Sloan School of Management.
Davis, F.D., 1989. Perceived usefulness, perceived ease of use, and user acceptance of
information technology. MIS Q. 13 (3), 319–340.
Davis, F.D., Bagozzi, R.P., Warshaw, P.R., 1989. User acceptance of computer technology: a
comparison of two theoretical models. Manag. Sci. 35 (8), 982–1003.
Davis, F.D., Bagozzi, R.P., Warshaw, P.R., 1992. Extrinsic and intrinsic motivation to use
computers in the workplace. J. Appl. Soc. Psychol. 22 (14), 1111–1132.
Demirkan, H., Spohrer, J., 2014. Developing a framework to improve virtual shopping in
digital malls with intelligent self-service systems. J. Retail. Consum. Serv. 21 (5),
860–868.
Dey, A., Sandor, C., 2014. Lessons learned: evaluating visualizations for occluded objects in
handheld augmented reality. Int. J. Hum. Comput. Stud. 72 (10–11), 704–716.
Domina, T., Lee, S.E., MacGillivray, M., 2012. Understanding factors affecting consumer
intention to shop in a virtual world. J. Retail. Consum. Serv. 19 (6), 613–620.
Ducoffe, R.H., 1996. Advertising value and advertising on the web. J. Adv. Res. 36 (5),
21–35.
Dumas, J., Fox, J., 2008. Usability testing: current practice and future directions. In: Sears,
A., Jacko, J. (Eds.), The Human-Computer Interaction Handbook. Fundamentals, Evolving
Technologies, and Emerging Application, Chap. 57. Associates (Taylor and Francis
Group), New York, pp. 1129–1149.
Fishbein, M., Ajzen, I., 1975. Belief, Attitude, Intention and Behavior: An Introduction to
Theory and Research. Addison Wesley, Reading, MA.
Fittkau & Maaß Consulting, 2013. Augmented reality im E-commerce retrieved September
25, 2016, http://www.fittkaumaass.de/news/augmented-reality-im-e-commerce.
Flanagan, J.C., 1954. The critical incident technique. Psychol. Bull. 51 (4), 327–358.
Fonseca, D., Martí, N., Redondo, E., Navarro, I., Sánchez, A., 2014. Relationship between
student profile, tool use, participation, and academic performance with the use of
augmented reality technology for visualized architecture models. Comput. Hum.
Behav. 31, 434–445.
Fornell, C., Larcker, D.F., 1981. Evaluating structural equation models with unobservable
variables and measurement error. J. Mark. Res. 18 (3), 39–50.
Friborg, O., Martinussen, M., Rosenvinge, J.H., 2006. Likert-based vs. semantic differential-
based scorings of positive psychological constructs: a psychometric comparison
of two versions of a scale measuring resilience. Personal. Individ. Differ. 40 (5),
873–884.
Gefen, D., Karahanna, E., Straub, D.W., 2003. Trust and TAM in online shopping: an
integrated model. MIS Q. 27 (1), 51–90.
Gervautz, M., Schmalstieg, D., 2012. Anywhere interfaces using handheld augmented
reality. IEEE Computer 45 (7), 26–31.
Google play, 2015. AUTO BILD augmented reality retrieved August 18, 2015, https://play.
google.com/store/apps/details?id=de.autobild.augmentedreality&hl=de.
Gurtner, S., Reinhardt, R., Soyez, K., 2014. Designing mobile business applications for
different age groups. Tech. Forcasting Soc. Chang. 88, 177–188.
Ha, S., Stoel, L., 2009. Consumer e-shopping acceptance: antecedents in a technology
acceptance model. J. Bus. Res. 62 (5), 565–571.
Hair, J.F., Ringle, C.M., Sarstedt, M., 2011. PLS-SEM: indeed a silver bullet. J. Mark. Theory
Pract. 19 (2), 139–152.
Hameed, M.A., Counsell, S., Swift, S., 2012. A conceptual model for the process of IT
innovation adoption in organizations. J. Eng. Technol. Manag. 29 (3), 358–390.
Hassenzahl, M., Tractinsky, N., 2006. User experience-a research agenda. Behav. Inform.
Technol. 25 (2), 91–97.
Hassenzahl, M., Diefenbach, S., Göritz, A., 2010. Needs, affect, and interactive product–
facets of user experience. Interact. Comput. 22 (5), 353–362.
Hausman, A.V., Siepke, J.S., 2009. The effect of web interface features on consumer online
purchase intentions. J. Bus. Ventur. 62 (1), 5–13.
Heise, D.R., 1970. The Semantic Differential and Attitude Research. In: Summers, G.F.
(Ed.), Attitude Measurement. Rand McNally, Chicago, IL, pp. 235–253.
Henseler, J., Ringle, C.M., Sinkovics, R.R., 2009. The use of partial least squares path
modeling in international marketing. Adv. Int. Mark. 20, 277–319.
Huang, T.-L., Liao, S., 2015. A model of acceptance of augmented-reality interactive
technology: the moderating role of cognitive innovativeness. Electron. Commer.
Res. 15 (2), 269–295.
Javornik, A., 2016a. ‘It's an illusion, but it looks real!’ consumer affective, cognitive and
behavioural responses to augmented reality applications. J. Mark. Manag. 1–25
http://dx.doi.org/10.1080/0267257X.2016.1174726.
Javornik, A., 2016b. Augmented reality: research agenda for studying the impact of its
media characteristics on consumer behaviour. J. Retail. Consum. Serv. 30, 252–261.
Javornik, A., Rogers, Y., Moutinho, A.M., Freeman, R., 2016. Revealing the shopper experience
of using a ‘magic mirror'augmented reality make-up application. Proceedings of the
2016 ACM Conference on Designing Interactive Systems, pp. 871–882.
Karahanna, E., Straub, D.W., Chervany, N.L., 1999. Information technology adoption across
time: a cross-sectional comparison of pre-adoption and post-adoption beliefs. MIS Q.
23 (2), 183–213.
Keil, M., Beranek, P., Konsynski, B., 1995. Usefulness and ease of use: field study evidence
regarding task consideration. Decis. Support. Syst. 13 (1), 75–91.
Kell, M., 2011. Augmented reality eyewear retrieved September 25, 2016, https://www.
mivision.com.au/augmenting-reality-eyewear/.
Kim, J., Forsythe, S., 2008. Adoption of virtual try-on technology for online apparel
shopping. J. Int. Mark. 22 (2), 45–59.
Please cite this article as: Rese, A., et al., How augmented reality apps are accepted by consumers: A comparative analysis using scales and
opinions, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.10.010
13
Kim, H.C., Hyun, M.Y., 2016. Predicting the use of smartphone-based augmented reality
(AR): does telepresence really help? Comput. Hum. Behav. 59, 28–38.
Lee, K.C., Chung, N., 2008. Empirical analysis of consumer reaction to the virtual reality
shopping mall. Comput. Hum. Behav. 24 (1), 88–104.
Lee, Y., Kozar, K.A., Larsen, K.R., 2003. The technology acceptance model: past, present,
and future. Commun. Assoc. Inf. Syst. 12 (1), 752–780.
Lee, H.H., Fiore, A.M., Kim, J., 2006. The role of the technology acceptance model in
explaining effects of image interactivity technology on consumer responses. Int.
J. Retail Distrib. Manag. 34 (8), 621–644.
Lee, H., Chung, N., Jung, T., 2015. Examining the cultural differences in acceptance
of mobile augmented reality: comparison of South Korea and Ireland. In:
Tussyadiah, I., Inversini, A. (Eds.), Information and Communication Technologies
in Tourism 2015. Springer, Cham Heidelberg New York Dordrecht London,
pp. 477–491.
Legris, P., Ingham, J., Collerette, P., 2003. Why do people use information technology?
A critical review of the technology acceptance model. Inf. Manag. 40 (3),
191–204.
Mashable (2012). Ikea adds augmented reality to 2013 catalog, retrieved September 25,
2016, www.mashable.com/2012/07/19/ikea-augmented-reality-catalog
Mekni, M., Lemieux, A., 2014. Augmented reality: applications, challenges and
future trends. In: Zaharim, A., Sopian, K., Psarris, K., Margenstern, M. (Eds.),
Applied Computational Science, Proceedings of the 13th International Conference on
Applied Computer and Applied Computational Science (ACACOS’14). WSEAS press,
pp. 205–214.
Nunnally, J.C., 1978. Psychometric Theory. 2nd ed. McGraw Hill, New York, NY.
Oh, H., Yoon, S.Y., Shyu, C.R., 2008. How can virtual reality reshape furniture retailing?
Cloth. Text. Res. J. 26 (2), 143–163.
Olsson, T., Lagerstam, E., Kärkkäinen, T., Väänänen-Vainio-Mattila, K., 2013. Expected user
experience of mobile augmented reality services: a user study in the context of
shopping centres. Pers. Ubiquit. Comput. 17 (2), 287–304.
Osgood, C.E., Suci, G.J., Tannenbaum, P.H., 1957. The Measurement of Meaning. University
of Illinois Press, Urbana, I.L.
Pan, Y., Zhang, J.Q., 2011. Born unequal: a study of the helpfulness of user-generated
product reviews. J. Retail. 87 (4), 598–612.
Pantano, E., 2014. Innovation drivers in retail industry. Int. J. Inf. Manag. 34 (3), 344–350.
Pantano, E., Di Pietro, L., 2012. Understanding consumer's acceptance of technology-based
innovations in retailing. J. Technol. Manag. Innov. 7 (4), 1–19.
Pantano, E., Naccarato, G., 2010. Entertainment in retailing: the influences of advanced
technologies. J. Retail. Consum. Serv. 17 (3), 200–204.
Pantano, E., Servidio, R., 2012. Modeling innovative points of sales through virtual and
immersive technologies. J. Retail. Consum. Serv. 19 (3), 279–286.
Pantano, E., Timmermans, H., 2014. What is smart for retailing? Procedia Environ. Sci. 22,
101–107.
Pohlmann, J., 2013. Virtuelle anprobe par excellence – brillenkauf bei mister
spex, usabilityblog retrieved September 25, 2016, http://www.usabilityblog.de/
2013/02/virtuelle-anprobe-par-excellence-%E2%80%93-brillenkauf-bei-mister-
spex/.
Poncin, I., Mimoun, M.S.B., 2014. The impact of “e-atmospherics” on physical stores.
J. Retail. Consum. Serv. 21 (5), 851–859.
Porter, C.E., Donthu, N., 2006. Using the technology acceptance model to explain how
attitudes determine internet usage: the role of perceived access barriers and demo-
graphics. J. Bus. Res. 59 (9), 999–1007.
Pousttchi, K., Hufenbach, Y., 2014. Engineering the value network of the customer
interface and marketing in the data-rich retail environment. Int. J. Electron. Commer.
18 (4), 17–42.
Radley, T., 2014. Window-framed!: virtual mirror becomes a RayBan reality retrieved
September 25, 2016, http://www.vm-unleashed.com/window-framed-virtual-
mirror-becomes-a-rayban-reality/.
Rauschnabel, P.A., Ro, Y.K., 2016. Augmented reality smart glasses: an investigation of
technology acceptance drivers. Int. J. Technol. Mark. 11 (2), 123–148.
Rekimoto, J., Nagao, K., 1995. The world through the computer: computer augmented
interaction with real world environments. In: Robinson, G. (Ed.), Proceedings of the
8th Annual ACM Symposium on User Interface and Software Technology. ACM,
New York, NY, pp. 29–36.
Segars, A.H., Grover, V., 1993. Re-examining perceived ease of use and usefulness: a
confirmatory factor analysis. MIS Q. 17 (4), 517–525.
Spreer, P., Kallweit, K., 2014. Augmented reality in retail: assessing the acceptance and
potential for multimedia product presentation at the PoS. Trans. Mark. Res. 1 (1),
20–35.
Stoyanova, J., Brito, P.Q., Georgieva, P., Milanova, M., 2015. Comparison of Consumer
Purchase Intention between Interactive and Augmented Reality Shopping Platforms
through Statistical Analyses. IEEE International Symposium on Innovations in
Intelligent SysTems and Applications (INISTA), pp. 1–8.
Subramanian, G.H., 1994. A replication of perceived usefulness and perceived ease of use
measurement. Decis. Sci. 25 (5–6), 863–874.
Sun, H., Zhang, P., 2008. An exploration of affect factors and their role in user technology
acceptance: mediation and causality. J. Am. Soc. Inf. Sci. Technol. 59 (8),
1252–1263.
Sutherland, I.E., 1964. Sketch pad a man-machine graphical communication system. In:
Pistilli, P.O. (Ed.), Proceedings of the SHARE Design Automation Workshop, DAC ‘64.
ACM, New York, NY, pp. 329–346.
Turner, M., Kitchenham, B., Brereton, P., Charters, S., Budgen, D., 2010. Does the technology
acceptance model predict actual use? A systematic literature review. Inf. Softw.
Technol. 52, 463–479.
Van der Heijden, H., 2004. User acceptance of hedonic information systems. MIS Q. 28 (4),
695–704.
14 A. Rese et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx

Van Krevelen, D.W.F., Poelman, R., 2010. A survey of augmented reality technologies,
applications and limitations. Int. J. Virtual Real. 9 (2), 1–20.
Venkatesh, V., 2000. Determinants of perceived ease of use: integrating control, intrinsic
motivation, and emotion into the technology acceptance model. Inf. Syst. Res. 11 (4),
342–365.
Venkatesh, V., Bala, H., 2008. Technology acceptance model 3 and a research agenda on
interventions. Decis. Sci. 39 (2), 273–315.
Venkatesh, V., Davis, F.D., 2000. A theoretical extension of the technology acceptance
model: four longitudinal field studies. Manag. Sci. 46 (2), 186–204.
Venkatesh, V., Morris, M.G., Davis, G.B., Davis, F.D., 2003. User acceptance of information
technology: toward a unified view. MIS Q. 27 (3), 425–478.
Wixom, B.H., Todd, P.A., 2005. A theoretical integration of user satisfaction and technology
acceptance. Inf. Syst. Res. 16 (1), 85–102.
Wu, J., Lederer, A., 2009. A meta-analysis of the role of environment-based voluntariness
in information technology acceptance. MIS Q. 33 (2), 419–432.
Yousafzai, S.Y., Foxall, G.R., Pallister, J.G., 2007a. Technology acceptance: a meta-analysis
of the TAM: part 1. J. Model. Manag. 2 (3), 251–280.
Yousafzai, S.Y., Foxall, G.R., Pallister, J.G., 2007b. Technology acceptance: a meta-analysis
of the TAM: part 2. J. Model. Manag. 2 (3), 281–304.
Zaichkowsky, J.L., 1985. Measuring the involvement construct. J. Consum. Res. 12 (3),
341–352.
ZVA (Zentralverband der Augenoptiker und Optometristen - Association of
Optometrists), 2016A. Augenoptik in Zahlen Branchenbericht 2015/16. ZVA
gefoerdert durch Bundesministerium fuer Wirtschaft und Energie retrieved
September 25, 2016, http://www.zva.de/branchenberichte.
Daniel Baier is Professor of Innovation and Dialogue Marketing at the University of
Bayreuth, Germany. His research interests are in market-oriented design of products
and services, adoption and diffusion of new products and services, multivariate statistics,
data and web mining. He studied computer science at the University of Karlsruhe where
he also completed his dissertation and habilitation in business administration. He has
published in journals like Advances in Data Analysis and Classification, Annals of Operations
Research, Creativity and Innovation Management, Journal of Econometrics, Journal of Retailing
and Consumer Services, and R&D Management.
Andreas Geyer-Schulz is Professor of Information Services and Electronic Markets
(Schroff endowed chair) at the Department of Economics and Business Engineering and
is co-founder and director of the Institute of Information Systems and Marketing at the
Karlsruhe Institute of Technology (KIT), Germany. His research focuses on the design,
analysis and implementation of innovative information and market services, especially
recommender systems and personalized information systems in the CRM context. He
has published in journals like Journal of Universal Computer Science, Journal of Systemics,
Cybernetics and Informatics, and Marketing ZFP.
Stefanie Schreiber completed her Master of Science degree in Business Administration at
Brandenburg University of Technology Cottbus-Senftenberg, Germany, and now works as
Research Assistant at the Chair of Marketing and Innovation Management. Her PhD thesis
focuses on consumer's acceptance of augmented reality technologies. She has published in
journals like Marketing ZFP - Journal of Research and Management and Journal of Retailing
and Consumer Services.

Alexandra Rese is Assistant Professor at the Chair of Innovation and Dialogue Marketing at
the University of Bayreuth, Germany. She completed her dissertation in sociology and
entrepreneurship at the University of Karlsruhe and her habilitation in business
administration at Brandenburg University of Technology Cottbus-Senftenberg. Her works
have appeared in journals like R&D Management, Creativity and Innovation Management,
International Journal of Automotive Technology and Management, and Journal of Retailing
and Consumer Services. Her current research focuses on the acceptance of innovative
applications in retailing, e.g. augmented reality, as well as abilities and roles in innovation
management.

Please cite this article as: Rese, A., et al., How augmented reality apps are accepted by consumers: A comparative analysis using scales and
opinions, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.10.010